13/11/2020
New issue alert, issue 9 is now available online: https://www.tandfonline.com/toc/iwbp20/21/9?nav=tocList
Attention Deficit Hyperactivity disorder (ADHD) is characterised by inattention, hyperactivity, and impulsivity, and found in 5–10% of children and adolescents. Severe cases often also have learning difficulties and/or autism. ADHD also increases the risk of many other problems including Oppositional Defiant Disorder/Conduct Disorder (ODD/CD), anxiety states, substance misuse, crime and violence. The severity of ADHD and these risks are reduced by treatment with stimulant medication.
All too often ADHD is under-diagnosed, under-treated and sets young people on an adverse developmental trajectory that blights their lives. A diagnostic test or a means of predicting treatment response could go a long way to avoiding this. Several papers in this edition of the WJBP shed light on the pathophysiology of ADHD and related conditions. Taken together with other recent contributions in this journal and others, it is clear that progress is being made in elucidating their underlying neurobiology, but much less clear how that will translate into benefits for patients.
Brain structure and function
As part of the Enhancing NeuroImaging and Genetics through Meta-Analysis (ENIGMA) consortium, Hoogman et al. (2017) combined data from 1713 participants with ADHD and 1529 controls from 23 sites (median age 14, range 4–63 years) and found reduced volumes in ADHD in several brain regions, including the amygdala (Cohen’s d = –0.19), accumbens (d = –0.15), putamen (d = –0.14), caudate (d = –0.11), hippocampus (d = –0.11), and intracranial volume (d = –0.10). The effect sizes were highest in children, but symptom scores or stimulant medication use did not influence results. Nor did the presence of comorbid psychiatric disorders.
Boedhoe et al. (2020) recently took this ENIGMA work forward by directly comparing ADHD, autism and Obsessive-Compulsive Disorder (OCD). Structural T1-weighted whole-brain MRI data were analysed in each site, using Freesurfer software and standardised processing protocols, to pool measures from each cohort without the need for raw data transfer. The data set comprised no fewer than 5827 healthy controls, 2271 patients with ADHD, 1777 with autism and 2323 with OCD, from 151 cohorts worldwide. Interesting trends emerged but no significant differences between disorders survived correction for multiple comparisons.
It should however be noted that neuroimaging studies and reviews using different methods may arrive at different results. Voxel-based morphometry (VBM), for example, measures density or volume at a voxel level and meta-analyses tend to use a ‘signed differential mapping’ procedure to pool study results (which increases power but can be prone to false positives). One such meta-analysis of structural and functional MRI studies showed, for example, that ADHD and OCD have possibly disorder-specific structural and functional alterations in basal ganglia and insula – reductions in ADHD but increases in OCD relative to controls. The same was true in frontal regions, in that rostro-dorsal medial frontal regions were decreased in structure and function in OCD but ventrolateral prefrontal regions were under-functioning in ADHD (Norman et al. 2016).
What is missing, however, is a clear demonstration of how these imaging features map on to key cognitive characteristics or symptoms of ADHD or other developmental disorders (Philip et al. 2012). There is some evidence that structural and functional abnormalities in the amygdala are just as evident, if not more so, in ODD/CD as compared to ADHD, and that the most consistent clinical association is with ODD/CD symptoms (Noordermeer et al. 2016). Comparing studies of the the two groups of patients, Puiu et al. (2018) found that they share similar cortico-limbic structural and functional alterations but differential associations. In ADHD, Impulsive Aggression (IA) was associated with prefrontal cortex (PFC) and cingulate activity, while Response Inhibition (RI) deficits were linked to hypoactivity in the dorso/ventro-lateral PFC, insula, and striatum. Across disorders, cortico-limbic dysfunctions underlied IA, while RI was mostly associated with aberrant PFC activity. Importantly, there is fairly consistent evidence that methylphenidate normalises these patterns in ADHD. A fMRI meta-analysis showed that methylphenidate significantly enhanced activation in bilateral inferior frontal cortex/insula during inhibition and time discrimination but had no effect on working memory networks (14 fMRI datasets, 212 children with ADHD). A more lenient threshold also revealed increased putamen activation (Rubia et al. 2014).
Focussing on the amygdala in ADHD, Van Dessel et al (2019) build on their previous findings with fMRI that amygdala dysfunction is associated with delay aversion in ADHD. In a careful combined structural and functional MRI study of 28 adolescents aged 10–18 with ADHD (three with ODD, none with CD) and 32 matched controls, reduced amygdala volumes (maximal in the area responsible for emotional processing) were associated with and indeed mediated delay related amygdala hyperactivity (maximal basolaterally, involved in stress response) with both objective delay-related stimuli processing and subjective delay aversion. Most of the patients (24) were medicated but withdrawn from treatment 72 hours before scanning, making the effects of medication unclear. Symptom severity was not examined but one wonders if the delay aversion described is related to emotional liability, stress intolerance or impulsivity. Perhaps there is an ADHD subgroup with pronounced amygdala disruption at high risk of aggression and violence.
Connectivity and correlates
It is of course clear that most thoughts and actions require co-ordinated activity across neural systems, but methodological complications and variation between labs has hindered the development of consensus. Diffusion (Tensor) Imaging (DTI) is the standard approach to examining structural connectivity in the brain, but voxel-based and tract-based analyses have delivered little agreement as to which regions are most affected. There is a general tendency to finding reduced connectivity in ADHD but even this is inconsistent in studies which have gone to the trouble to confirm that motion in the scanner could not account for the results (Aoki et al. 2018).
Resting-state fMRI is the most commonly used approach to study functional connectivity in the brain. The default-mode network (DMN) reflects a set of interacting brain regions (i.e. medial prefrontal cortex, precuneus, posterior cingulate cortex (PCC), and medial temporal regions) with coherent neural activation when internal thoughts are self-generated, and higher activity at rest than when performing externally focussed goal-directed tasks (Raichle et al. 2001).
Sutcubasi et al. (2020) have conducted a very useful meta-analysis of 20 seed based resting state functional connectivity studies in 944 (mainly treated) patients with ADHD and 1121 controls. They found reduced connectivity within the core DMN sub-system (from a PCC seed), alongside increases in a broad set of nodes in systems involved in cognition and motivation, especially in children and adolescents. DMN overactivity at rest and/or underactivity on activation is however a common observation in many psychiatric disorders, but DMN dysfunction could have differential clinical associations – such as with mind-wandering and inattention in ADHD, and perhaps anxious rumination in OCD.
Li et al. (2020) show DMN overactivity in OCD (N = 42, mean age 27 years) compared to healthy controls, with increased hippocampal connectivity to fronto-temporal regions showing different sub-regional associations with obsessive, compulsive and anxious symptom severity scores. Hippocampal functional connectivity to left middle temporal gyrus correlated with obsessions and anxiety, and a mediation analysis showed that this mediated the relationship between obsessions and anxiety. This could perhaps reflect a difficulty in using the regularities of previous experience to regulate anxiety.
Almost all ADHD treatment studies suggest that methyphenidate normalises DMN circuits (Santos et al. 2019), just as many treatments in psychiatry may also do. Woelfer et al. (2019) conducted a randomised placebo-controlled study of the effects of ketamine on functional connectivity in 53 healthy participants. They found that Brain-Derived Neurotrophic Factor (BDNF) plasma levels increased and seemed to drive ketamine induced decreases in resting-state functional connectivity between PCC and ventro-medial PFC. BDNF is a neurotrophin involved in neurogenesis and synaptic plasticity, especially in the hippocampus, and has been implicated in the pathophysiology of several neuropsychiatric disorders. Plasma BDNF and its common genetic polymorphisms are not consistently abnormal in ADHD but some specific haplotypes or polymorphisms may be (Bonvicini et al. 2018). Given that ketamine increases synaptic plasticity, and has recently been licenced for the treatment of treatment resistant depression, its mode of action may include normalising resting state functional connectivity and this may be a way of targeting its use in depression and other psychiatric disorders.
Conclusions
ADHD, autism, ODD/CD and OCD are common neurodevelopmental disorders that frequently co-occur but are usually distinguishable clinically. Given the overlap in genetic and environmental risk factors – some of which have known impact on brain imaging measures – some overlap in neuroimaging findings across the disorders is to be expected. There may not be specific neuroimaging phenotypes, but there may well be differential associations between disrupted anatomy and function, potentially mediating cognitive processes such as deficits in frontal inhibition and social cognition, and symptoms of the conditions. A transdiagnostic approach is worth pursuing alongside disorder specific studies, but ADHD has a specific treatment that can transform lives and has clear effects on functional MRI. A refined translational neuroimaging of ADHD and related developmental disorders will however require a concerted attempt to address these issues across multiple laboratories using compatible methodologies.
Frequency: Yearly ISSN: 1562-2975 eISSN: 1814-1412 https://www.tandfonline.com/doi/abs/10.1080/15622975.2020.1823694